基于FPGA的分子动力学模拟交互控制系统

摘要/Abstract

摘要： 在分子动力学模拟系统中，实现分子间短程力的计算需要频繁地传输与大量的粒子数据访问。为了减轻CPU的计算负载，可以使用FPGA加速计算。但是，在基于FPGA的分子动力学模拟系统中，短程力计算模块面临巨大的数据传输压力以及访存冲突问题。针对这些问题，基于FPGA上有限的硬件资源，提出一种交互控制系统。该系统由取数控制模块与粒子数据解析模块组成。整个系统通过合理的数据编排以及2个模块的协同工作，实现粒子数据从片上存储到短程力计算模块的快速可靠的传输。通过硬件仿真和板级实验验证了该系统在处理粒子数据过程中的有效性和可靠性。

关键词: 分子动力学, 短程力, 交互控制系统, 现场可编程门阵列, 并行计算

Abstract: In molecular dynamics simulation systems, the computation of range-limited forces between molecules requires frequent transfers and large amounts of particle data access. To reduce the computational load on the CPU, FPGAs can be used to accelerate the computation. However, in the FPGA-based molecular dynamics simulation system, the range-limited force calculation module faces huge data transfer pressure and access conflict problems. To address these problems, an interactive control system is designed based on the limited hardware resources on FPGAs. The system consists of a fetch control module and a particle data parsing module. The whole system achieves fast and reliable transmission of particle data from on-chip storage to the range-limited force calculation module through reasonable data arrangement and cooperative work of the two modules. The effectiveness and reliability of the system in processing particle data are further verified by hardware simulation and board-level experiments.

Key words: molecular dynamics, range-limited force, interactive control system, Field Programmable Gate Array （ FPGA）, parallel computing

王鑫, 吴俊辉, . 基于FPGA的分子动力学模拟交互控制系统[J]. 计算机与现代化, 2022, 0(09): 25-31.

WANG Xin, WU Jun-hui, . FPGA-based Interactive Control System for Molecular Dynamics Simulation[J]. Computer and Modernization, 2022, 0(09): 25-31.

参考文献

［1］殷开梁. 分子动力学模拟的若干基础应用和理论［D］. 杭州:浙江大学, 2006.
［2］ HOLLINGSWORTH S, DROR R. Molecular dynamics simulation for all［J］. Neuron, 2018,99（6）:1129-1143.
［3］ KARPLUS M, MCCAMMON J. Molecular dynamics simulations of biomolecules［J］. Nature Structural Biology, 2002,9（9）:646-652.
［4］ HOSPITAL A, GOI J R, OROZCO M, et al. Molecular dynamics simulations: Advances and applications［J］. Advances and Applications in Bioinformatics and Chemistry, 2015,8:37-47.
［5］ WANG H Q, PENG S L, ZHU X Q, et al. A method to accelerate GROMACS in offload mode on Tianhe-2 supercomputer［C］// 2015 15th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing. 2015:781-784.
［6］ CHIU M, KHAN M A, HERBORDT M C. Efficient calculation of pairwise nonbonded forces［C］// 2011 IEEE 19th Annual International Symposium on Field-Programmable Custom Computing Machines. 2011:73-76.
［7］ WU C S, GENG T, YANG C, et al. A communication-efficient multi-chip design for range-limited molecular dynamics［C］// 2020 IEEE High Performance Extreme Computing Conference （HPEC）. 2020:1-8.
［8］ BROWN W M, WANG P, PLIMPTONS J, et al. Implementing molecular dynamics on hybrid high performance computers-short range forces［J］. Computer Physics Communications, 2011,182（4）:898-911.
［9］ GLASER J, NGUYEN T D, ANDERSON J A, et al. Strong scaling of general-purpose molecular dynamics simulations on GPUs［J］. Computer Physics Communications, 2015,192:97-107.
［10］ALAM S R, AGARWAL P K, SMITH M C, et al. Using FPGA devices to accelerate biomolecular simulations［J］. Computer, 2007,40（3）:66-73.
［11］郭禾,龙珠,王宇新,等. 基于FPGA的分子动力学模拟系统设计［J］. 微电子学与计算机, 2009,26（8）:246-248.
［12］YANG C, GENG T, WANG T Q, et al. Fully integrated FPGA molecular dynamics simulations［C］// Proceedings of the 2019 International Conference for High Performance Computing, Networking, Storage and Analysis. 2019:1-31.
［13］GU Y F, VANCOURT T, HERBORDT M C. Accelerating molecular dynamics simulations with configurable circuits［J］. IEE Proceedings-Computers and Digital Techniques, 2006,153（3）:189-195.
［14］SHAW D E, GROSSMAN J P, BANK J A, et al. Anton 2: Raising the bar for performance and programmability in a special-purpose molecular dynamics supercomputer［C］// Proceedings of the 2014 International Conference for High Performance Computing, Networking, Storage and Analysis. 2014:41-53.
［15］KASAP S, BENKRID K. Parallel processor design and implementation for molecular dynamics simulations on a FPGA-based supercomputer［J］. Journal of Computers, 2012,7（6）:1312-1328.
［16］王海强. 天河2号上CPU/MIC协同的分子动力学模拟软件GROMACS并行加速技术研究［D］. 长沙:国防科学技术大学, 2015.
［17］KASAP S, BENKRID K. A high performance implementation for molecular dynamics simulations on a FPGA supercomputer［C］// 2011 NASA/ESA Conference on Adaptive Hardware and Systems （AHS）. 2011:375-382.
［18］LARSON R H, SALMON J K, DRORR O, et al. High-throughput pairwise point interactions in Anton, a specialized machine for molecular dynamics simulation［C］// Proceedings of the 14th International Symposium on High Performance Computer Architecture. 2008:331-342.
［19］CHIU M, HERBORDT M C. Molecular dynamics simulations on high-performance reconfigurable computing systems［J］. ACM Transactions on Reconfigurable Technology and Systems （TRETS）, 2010,3（4）:1-37.
［20］WU C, GENG T, BANDARA S, et al. Upgrade of FPGA range-limited molecular dynamics to handle hundreds of processors［C］// Proceedings of the 29th IEEE International Symposium on Field-Programmable Custom Computing Machines （FCCM）. 2021:142-151.
［21］SHAW D E, DROR R O, SALMON J K, et al. Millisecond-scale molecular dynamics simulations on Anton［C］// Proceedings of the 2009 Conference on High Performance Computing Networking, Storage and Analysis. 2009:1-11.
［22］GROSSMAN J P, TOWLES B, GRESKAMP B, et al. Filtering, reductions and synchronization in the Anton 2 network［C］// Proceedings of the 29th International Parallel and Distributed Processing Symposium. 2015:860-870.
［23］CHIU M, HERBORDT M C. Towards production FPGA-accelerated molecular dynamics: Progress and challenges［C］// Proceedings of 2010 4th International Workshop on High Performance Reconfigurable Computing Technology and Applications （HPRCTA）. 2010:1-8.
［24］YAO Z, WANG J S, LIU G R, et al. Improved neighbor list algorithm in molecular simulations using cell decomposition and data sorting method［J］. Computer Physics Communi-cations, 2004,161（1-2）:27-35.
［25］SHAW D E. A fast, scalable method for the parallel evaluation of distance-limited pairwise particle interactions［J］. Journal of Computational Chemistry, 2005,26（13）:1318-1328.
［26］ANDOH Y, YOSHII N, FUJIMOTO K, et al. MODYLAS: A highly parallelized general-purpose molecular dynamics simulation program for large-scale systems with long-range forces calculated by fast multipole method （FMM） and highly scalable fine-grained new parallel processing algorithms［J］. Journal of Chemical Theory and Computation, 2013,9（7）:3201-3209.
［27］KHAN M A, CHIU M, HERBORDT M C. FPGA-accelerated molecular dynamics［M］// High-Performance Computing Using FPGAs. Springer, 2013:105-135.
［28］YANG C,GENG T, WANG T Q, et al. Molecular dynamics range-limited force evaluation optimized for FPGAs［C］// Proceedings of the 30th International Conference on Application-specific Systems, Architectures and Processors （ASAP）. 2019:263-271.

[1]	谭金林, 范文童, 刘亚虎, 梁志锋, 王梁, 刘斌, 黄斌. 基于软硬件协同加速框架的遥感图像目标检测[J]. 计算机与现代化, 2022, 0(06): 109-115.
[2]	项武铭，李雪巍. 基于CUDA的梯级泵站调度算法实现[J]. 计算机与现代化, 2018, 0(11): 60-.
[3]	赵博颖，肖鹏，张力. 基于Docker的MPI和OpenMP混合编程[J]. 计算机与现代化, 2018, 0(05): 60-.
[4]	桑喆1，邓川1，苟聪1，刘开兴2，白明泽1. 基于JNI和C+〖KG-*3〗+的Intel集成众核并行方法[J]. 计算机与现代化, 2018, 0(04): 32-.
[5]	邬可可,黄国伟,孔令晶. 一种灵活的椭圆曲线密码并行化方法[J]. 计算机与现代化, 2018, 0(02): 71-.
[6]	王龙，姚文明. 基于Spark的并行遗传算法在物流配送问题中的应用[J]. 计算机与现代化, 2018, 0(01): 19-22.
[7]	栗家棋，郑昊. 基于矢量地图数据的异构混合快速绘制框架[J]. 计算机与现代化, 2018, 0(01): 56-61.
[8]	孙立华1，胡牧1，孟庆强1，钱亚康1，王松2. 配网线损大数据高性能计算解决方案[J]. 计算机与现代化, 2016, 0(12): 42-46,50.
[9]	汪方良，施慧彬. 基于OpenCL的人脸识别与隐藏系统[J]. 计算机与现代化, 2016, 0(1): 16-19,58.
[10]	赵成龙1，施慧彬1，俞忻峰2. 基于OpenCL的双GPU基数排序算法[J]. 计算机与现代化, 2015, 0(1): 27-30.
[11]	贺维维，吴京，曾瑶源. 基于MPI+OpenMP的红外弱小目标检测并行计算[J]. 计算机与现代化, 2014, 0(7): 53-57.
[12]	王兴，王介君，孙宁，汪瑶. 基于GPU的交叉相关外推并行化算法[J]. 计算机与现代化, 2014, 0(2): 213-218.
[13]	张唯唯，张玉洁. 基于GPU的并行报文分类方法[J]. 计算机与现代化, 2014, 0(11): 9-14.
[14]	刘晴和，杨云，贺兴亚，徐文春，周媛媛. 基于业务分割的并行式大数据迁移策略研究[J]. 计算机与现代化, 2014, 0(11): 86-89.
[15]	付康. 甚高频多普勒雷达的数字前端设计[J]. 计算机与现代化, 2013, 1(9): 133-136.